Medical knowledge extraction has great potential to improve the treatment quality of hospitals. In this paper, we propose a clinical problem-action relation extraction method. It is based on clinical semantic units and event causality patterns in order to present a chronological view of a patient's problem and a physician's action. Based on our observation, a clinical semantic unit is defined as a conceptual medical knowledge for a problem and/or action. Since a clinical event is a basic concept of the problem-action relation, events are detected from clinical texts based on conditional random fields. A clinical semantic unit is segmented from a sentence based on time expressions and inherent structure of events. Then, a clinical semantic unit is classified into a problem and/or action relation based on event causality features in support vector machines. The experimental result on Korean medical collection shows 78.8% in F-measure when given the answer of clinical events. This result shows that the proposed method is effective for extracting clinical problem-action relations.

Social media such as Twitter has come to reflect the reaction of the general public to major events. Since posts are short and noisy, it is hard to extract reliable events based on word frequency. Even though an event term appears in a particularly low frequency, as long as at least one reliable user mentions the term, it should be extracted. This paper proposes an event extraction method which combines user reliability and timeline analysis. The Latent Dirichlet Allocation (LDA) topic model is adapted with the weights of event terms on timeline and reliable users to extract social events. The reliable users are detected on Twitter according to their tweeting behaviors: socially well-known users and active users. Reliable and low-frequency events can be detected based on reliable users In order to see the effectiveness of the proposed method, experiments are conducted on a Korean tweet collection; the proposed model achieved 72% in precision. This shows that the LDA with timeline and reliable users is effective for extracting events on the Twitter test collection.

In this paper, the authors propose a novel bias detection method based on social preference learning for targets on competing topics such as “GalaxyTab vs. iPad” in Twitter. People tend to evaluate a topic by expressing their opinions towards the associated targets such as price and quality. To exploit characteristics of social data, targets are extracted by a modified HITS algorithm on a tripartite graph. The main contribution is that social preferences are learned with explicit sentiment, latent sentiment as social semantics, and lexical sentiment as contextual semantics on targets of the topic, and that the individual preference is considered together with social preferences for the bias detection of a tweet. Experimental results on Twitter collection show significant improvements over all baseline methods. The results indicate that the method deals with not only the lack of a sentiment lexicon but also social and contextual semantics on targets of social users.

Typical pseudo-relevance feedback methods assume the topretrieved documents are relevant and use these pseudo-relevant documents to expand terms. The initial retrieval set can, however, contain a great deal of noise. In this paper, we present a clusterbased resampling method to select better pseudo-relevant documents based on the relevance model. The main idea is to use document clusters to find dominant documents for the initial retrieval set, and to repeatedly feed the documents to emphasize the core topics of a query. Experimental results on large-scale web TREC collections show significant improvements over the relevance model. For justification of the resampling approach, we examine relevance density of feedback documents. A higher relevance density will result in greater retrieval accuracy, ultimately approaching true relevance feedback. The resampling approach shows higher relevance density than the baseline relevance model on all collections, resulting in better retrieval accuracy in pseudo-relevance feedback. This result indicates that the proposed method is effective for pseudo-relevance feedback.

In recent years, statistical approaches on ATR (Automatic Term Recognition) have achieved good results. However, there are scopes to improve the performance in extracting terms still further. For example, domain dictionaries can improve the performance in ATR. This paper focuses on a method for extracting terms using a dictionary hierarchy. Our method produces relatively good results for this task.